There has been a lot of discussion about the way in which massive open online courses (MOOCs) might change the landscape of education. A transformation is likely, but in a form more nuanced than current conversations imply (see go.nature.com/cynqvp).

Another question has had less attention: how will MOOCs affect research and scholarly enterprise at universities? Will the courses lead to a decline in the amount or quality of study? Or will a huge increase in online learning improve research?

We are both deeply involved in the MOOC culture at the Massachusetts Institute of Technology (MIT) in Cambridge. We are members of the university's Task Force on the Future of MIT Education, which focuses on the use and effect of MOOCs, and we both run active research groups, so we appreciate the dual tugs of teaching and scholarship. Our opinion is that, at least for residential, research-focused universities, the main impact of MOOCs on research culture will be to aid interdisciplinary work by enabling greater access to conceptual or technical tuition.

Credit: ILLUSTRATION BY CHRIS RYAN/NATURE

Many factors influence scholarship and research in a university. These include funding, space, equipment, research collaborations, the number of faculty members, the presence of faculty members on campus and the time available for research. Research is often deeply affected by the insight and questioning of smart undergraduate students or postdocs. Fields move forwards through the ability to think in novel, cross-disciplinary ways, and through familiarity with state-of-the-art techniques.

Which parts of that equation will MOOCs affect? A common view is that they will replace professors. Clearly, this would diminish the volume of research and stifle the collegial discourse that fosters collaboration and original thinking. We do not consider this to be a likely outcome — except perhaps in parts of the world where students have limited access to quality education and where a MOOC might stand in for face-to-face teaching, possibly leading to a cut in faculty numbers. Data on this point are not yet available owing to the newness of these teaching tools.

In research-intensive universities, it is our experience that the time in a faculty member's schedule that is freed up through the use of online material is invested back into 'flipped' classrooms — students have already studied lecture material online, meaning that class time is used for discussion, hands-on activities, problem solving and research. Similarly, cutting-edge graduate classes thrive on faculty–student interaction. Any reduction in lecture time is reinvested in the classroom to make the experience more comprehensive and fruitful for graduate students.

DIGITAL LEARNING A Nature and Scientific American special issue nature.com/digitallearning

When online learning tools were used to teach electrical engineering to MIT undergraduates, faculty time that would have been devoted to lecturing went back into the course as alternative ways of teaching the material. These included practical demonstrations, hands-on experiments and interactive laboratory sessions featuring design, creation, debugging, small group discussions and one-to-one student–instructor feedback.

The increase in problem solving and discussion could expose more undergraduates to research questions and strategy earlier and more extensively than at present. This could encourage undergraduates to pursue research during their degree period or beyond, thereby enriching research culture.

For faculty members devising MOOCs and developing additional learning tools, there might be a short-term dip in research productivity. Such a dip is experienced by any devoted teacher who devises a new course, but the high-tech nature of MOOCs can involve a bigger time investment than normal. Our colleagues at MIT report that developing a new MOOC takes between half a semester and a full one, which lasts about 14 weeks.

The most important impact on research will probably be interdisciplinary online education. Say that an MIT student needs to learn current techniques in machine learning to apply to her main area of study, mechanical engineering. An appropriate class is not offered at MIT until the next spring, but the University of California, Berkeley, offers an online course that she can take this summer. She takes the course and an adviser at MIT vets her progress. By the autumn, the student is ready to apply her newly acquired skills to her thesis research.

Online education is unlikely to dent faculty engagement in research.

A related and crucial benefit of MOOCs to research stems from modules that will help investigators to straddle fields by aiding fluid learning. Modules can range from, for example, a 10-minute video segment to the content of multiple full-length lectures. Thus, for example, a postdoctoral neuroscientist might need to learn how to measure changes in cell shape. A full course is not necessary. Multiple modules explaining relevant concepts are available online from several universities. The postdoc uses these to increase his competency faster, more cheaply and more comprehensively than he would by taking a course, visiting a relevant laboratory or attending a meeting. This sort of learning has the potential to improve research output.

The serious decline of research funding in the United States and in other parts of the world is reducing the ability of researchers to travel to acquire training in person. Although we in no way suggest that virtual education can entirely plug this gap, it can help scientists to gain some of the conceptual and technical expertise that is necessary to advance research.

Although we feel that online education is unlikely to dent faculty engagement in research, it will be important to monitor carefully the amount of time that faculty members spend on developing MOOCs. We suggest that this is done at the departmental level to ensure that a push for online educational development is balanced by responsiveness to any research challenges that result.